HYDROGEN-SENSITIVE GaN SCHOTTKY DIODES

 

Jihyun Kim(1), B. P. Gila(2), G. Y. Chung(3), C. R. Abernathy(2), S. J. Pearton(2) and F. Ren(1)

 

(1)   Department of Chemical Engineering, University of Florida, Gainesville FL 32611 USA

(2)   Department of Materials Science and Engineering University of Florida, Gainesville FL 32611

(3)   Sterling Semiconductor, Tampa FL 33619 USA

 

 ABSTRACT

 

 Pt/n-GaN and Pd/n-GaN Schottky diodes were characterized for their response to hydrogen as a function of measurement temperature. Even at modest temperatures(80~140°C), an 80mm diameter diode shows a large increase(³0.5mA) in forward current upon introduction of ~10% H2 into a N2 ambient. The change in current increases with measurement temperature and begins essentially immediately upon introduction of the hydrogen. Cycling the ambient from N2 to 10%H2 in N2 and back to N2 produces reproducible cycling of the forward current at fixed forward bias. The decrease in barrier height of Pt on GaN was 50mV at 25°C and 70mV at 150°C upon introduction of H2 into the ambient,with lower values for Pd. At high temperature, the time response of the sensors appears to be controlled by hydrogen diffusion to the metal/GaN interface, while at low temperatures(<100°C), dissociation of the gas appears to be the rate-determining step.

INTRODUCTION

 There has long been an interest in developing solid-state gas sensors(1-6). It was recognized early that the threshold voltage of field effect transistors or the barrier height of Schottky diodes fabricated in Si, SiC, ZnO and the semiconductors was altered upon exposure to hydrogen-containing gases, leading to changes in current in these structures for a given set of biasing conditions(1-7). In the intervening  period there has been a steady improvement in the understanding of the hydrogen-sensing mechanism(8-15). There is particular interest in the development of wide bandgap semiconductor gas sensors because of their potential for high temperature operation and the ability to integrate them with power or microwave electrodes or with UV solar-blind detectors and emitters fabricated in the same materials. There have been a number of reports on the gas-sensing properties of GaN Schottky diodes(16-20), but no comparisons of how different metals perform and little information in the temperature dependence of the time response. In this paper, we compare the characteristics of Pd/GaN and Pt/GaN Schottky diodes for sensing dilute H2 concentrations in N2 ambients at different temperatures. The Pt metallization is found to produce large changes in forward current of the diodes due to its higher catalytic cracking efficiency for hydrogen. Both types of diodes are found to provide sensitive detection of hydrogen at temperatures as low as 25°C.

 

EXPERIMENTAL

 The starting sample were 6mm thick n-GaN(n~3´1017 cm-3) layers grown on Al2O3 substrates by Metal Organic Chemical Vapor Deposition. Front-side ohmic contacts of Ti/Al/Pt/Au were formed by lift-off and subsequent annealing at 600°C. This Schottky contacts were formed by lift-off of e-beam deposited Pd or Pt, 24nm and 15nm thick respectively. The contact diameter was 80mm in all cases. The devices were wire-bonded to a test fixture using Ti/Au bond-pads and Au wires for contact. Figure 1 shows a photograph of a completed device. The gas-sensing experiments were carried out in a system described elsewhere(21), over a range of temperatures(25~200°C) with either N2 or 10% H2 in N2 ambient.

 

RESULTS AND DISCUSSION

 Figure 2 shows the time response of Pt/GaN diodes to switching from N2 to 10% H2 in N2 and back again, at a measurement temperature of 110°C with the diode biased at 1.5V forward bias. Note the rapid change in forward current as the H2 is introduced and the shower decay back towards the original current level upon switching back to N2. The barrier height, fB is given by

fB=fm-cS

  =eVBI+d

where fm is the metal workfunction, cS the electron affinity of GaN, e is the electronic charge, d is the Fermi energy below the conduction band and VBI is the build-in voltage. From the forward I-V characteristics, we were able to measure DfB of 50mV at 25°C and 70mV at 150°C for Pt/GaN diodes upon exposure to the 10% H2 in N2 ambient. The lowering of the barrier due to accumulation of hydrogen at the Pt/GaN interface is consistent with the formation of the dipole layers at the interface, as suggested previously(17).

 Figure 3 shows the response of a Pt/GaN diode to repeated cycling of the ambient from N2 to 10% H2 in N2, with the sample held at 100°C. Even at this relatively modest temperature, the change in forward current for a fixed bias of 1.5V is readily detectable and shows there is sufficient cracking of the H2 for the diode to be a sensitive gas detector. Note also that there is highly reproducible saturation current for each of the gas ambient employed.

 The temperature dependence of the decay in forward current of 1.5V for Pt/GaN diodes is shown in Figure 4. The size of the change in current, and hence the change in fB, depends on temperature since cracking of the gas is more effective at higher temperatures. To gain some idea of the rate-determining mechanisms that are operative, we took the change in forward current at a fixed time of 50 seconds after switching of the gas and plotted this change in Arrhenius form(Figure 5).

 There are several features apparent in this data-firstly, the Pt/GaN diodes has a higher signal response than comparable Pd/GaN diodes. We ascribe this to the greater catalytic cracking efficiency of Pt for H2. Secondly, the data does not show a single slope, ie, there are multiple mechanisms at work. At high temperatures(³110°C), the activation energy is of order 0.70eV. This is consistent with the activation energy for diffusion of hydrogen in materials such as metals(22). This suggests that the time needed for the atomic hydrogen to diffuse through the Pt to the interface with the GaN is the rate-limiting step under these conditions. As the temperature is lowered below 100°C, there are obviously other mechanism at work and it is plausible that under these conditions the initial dissociation of the H2 is one of the rate-limiting steps, along with competing processes such as recombination of atomic hydrogen to form H2 molecules.

  Figure 6 shows forward I-V characteristics for Pd/GaN diodes at 170°C(top) or 200°C(bottom) in either N2 or 10% H2 in N2 ambients. The changes in forward current upon introducing the H2 are smaller than achieved with Pt/GaN diodes under the same conditions, as discussed above, but are still of sufficient magnitude to be useful as gas detectors at moderate temperatures. The choice of metallization may be dictated by the required operating temperature for the gas sensors. For example it may be necessary to use W-based schemes(23) or carbides that have extremely high thermal stability if operating temperatures above 400°C are required.

 The change in forward current for Pd/GaN diodes at two different temperatures are shown in Figure 7 for cycling between N2 and 10% H2 in N2. The forward current density J can be expressed as

 

where k is Boltzman’s constant, T is the absolute measurement temperature and J0 is dependent on fB. From the forward I-V characteristics, we found a change in barrier height of 30mV at 150°C and 60mV at 200°C upon introduction of H2 into the ambient for the Pd diodes.

 

SUMMARY AND CONCLUSIONS

 Both Pt/GaN and Pd/GaN Schottky diodes were examined for their temperature and time response to introduction of hydrogen into the ambient. Significant decreases in barrier height were observed in both cases, leading to increase in forward current. The Pt-GaN diodes show larger changes in current, due to more effective catalytic dissociation of H2 relative to use of a Pd contact. The GaN materials system appears to be very promising for use in combustion gas detection, especially as part  of integrated sensor structures that could also detect UV radiation.

 

ACKNOWLDEGEMENTS

 

 The work at UF is partially supported by NSF CTS 991173 and NASA(NAG10-316), monitored by Dr.William Knott.

References

  1. K. I. Lundström, M. S. Shivaraman, and C. M. Svensson, J. Appl. Phys., 46 3876(1975).
  2. I. Lundström, S. Shivaraman, C. Svensson, and L. Lundkvist, Appl. Phys. Lett. 26 55(1975).
  3. L. Stiblert and C. Svensson, Rev. Sci. Instrum. 46, 1206 (1975).
  4. M. C. Steele, J. W. Itile and B. A. MacIver, J. Appl. Phys. 47 2537(1976)
  5. I. Lundström, M. S. Shivaraman, L. Stiblert, and C. Svensson, Rev. Sci. Instrum. 47, 738 (1976)
  6. K. Ito, Surf. Sci. 86 345(1979).
  7. P. F. Ruths, S. Ashok, S. J. Foash and J. M. Ruths, IEEE Transaction on Electron. Dev. 28 1003(1981)
  8. A. L. Spetz, P. Tobias, L. Unéus, H. Svenningstorp, L.-G. Ekedahl and I. Lundström, Sens. Actuators B70 67(2000).
  9. G. W. Hunter, C.C. Liu, D. Makel, MEMS Handbook, ed. M. G. Hak (CRC Press, Boca Raton,2001).
  10. H. Svenningstorp, P. Tobias, I. Lundström, P. Salomonsson, P. Mårtensson, L.-G. Ekedahl and A. L. Spetz, B57 159(1999).
  11. G. W. Hunter, P. G. Neudeck, R. S. Okojie, G. M. Beheim, V. Thomas, L. Chen, D. Lukco, C. C. Liu, B. Ward and D. Makel, Proc. ECS Vol 01-02 212(2002)
  12. A. L. Spetz, A. Baranzahi, P. Tobias, and I. Lundström, Phys. Stat. Sol. A162 493(1997).
  13. E. J. Connolly, G. M. O'Halloran, H. T. M. Pham, P. M. Sarro and P. J. French, Sens. Actuators A99 25(2002)
  14. L. Chen, G. W. Hunter, and P. G. Neudeck, J. Vac. Sci. Technol A 15, 1228 (1997)
  15. P. Tobias, A. Baranzahi, A. L. Spetz, O. Kordina, E. Janzen and I. Lundstrom, IEEE Electron. Dev. Lett. 18 287(1997).
  16. J. Schalwig, G. Muler, M. Eickoff, O. Ambacher, and M. Stutzmann, Mat. Sci. Eng. B 101 86(2002).
  17. J. Schalwig, G. Muller, U. Karrer, M. Eickhoff, O. Ambacher, M. Stutzmann, L. Gorgens and G. Dollinger, Appl. Phys. Lett. 80 1222(2002).
  18. B. P. Luther, S. D. Wolter, S. E. Mohney, Sens. Actuators B56 164(1999).
  19. J. Schalwig, G. Muller, O. Ambacher and M. Stutzmann, Phys. Stat. Sol. A185 39(2001)
  20. R. Neuberger, G. Muler, O. Ambacher, and M. Stutzmann, Phys. Stat. Solidi A 183 R10(2001)
  21. Jihyun Kim, B. P. Gila, C. R. Abernathy, G. Y. Chung, F. Ren and S. J. Pearton, J. Electron. Mater.(in press)
  22. S. J. Pearton, J. W. Corbett, and Michael Stavola, Hydrogen in Crystalline Semiconductor(Springer-Verlag, Berlin 1991)
  23. S. J. Pearton, J. C. Zolper, R. J. Shul, F. Ren, J. Appl. Phys. 86,1 (1999)

Figure Captions

 

Figure 1. Photograph of completed Pt/n-GaN diode

 

Figure 2. Time response at 110°C of forward current of a Pt/GaN diode forward biased at 1.5V, upon switching from N2 ambient to 10% H2 in N2 and then back to N2.

 

Figure 3. Time response at 100°C of forward current of a Pt/GaN diode biased at 1.5V, upon cycling the ambient between pure N2 and 10% H2 in N2.

 

Figure 4. Decay of forward current as a function of measuremental temperature for Pt/GaN diodes at 1.5V forward bias after switching from 10% H2 in N2 ambient to pure N2.

 

Figure 5. Arrhenius plot of change in forward current of Pt/GaN diodes 50 secs after switching from 10% H2 in N2 ambient to pure N2.

 

Figure 6. Forward I-V characteristics from Pd/GaN diodes measured in N2 or 10% H2 in N2 ambient at either 170°C(top) or 200°C(bottom).

 

Figure 7. Forward current at 0.5V bias at 150°C(top) or 200°C(bottom) for Pd/GaN diodes cycled in N2 and 10% H2 in N2 ambients.

Figure 1.

 


Figure 2.

 

 
Figure 3.

 
Figure 4.

 
Figure 5.


Figure 6.

 

Top

Bottom

 
Figure 7.

 

Top

Bottom